INDEX

Explanations

numbers and numerical progressions

oai_token-act-pair · gpt-3.5-turbo

numerical values related to measurements and statistics

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ACTION

-0.65

rored

-0.61

owship

-0.61

AFP

-0.57

»Ĵ

-0.56

"))

-0.55

ALSE

-0.55

GC

-0.54

PASS

-0.54

yrinth

-0.53

POSITIVE LOGITS

 onwards

1.26

 onward

1.18

 standpoint

0.89

 beginnings

0.85

 inception

0.81

to

0.79

 perspective

0.74

 downwards

0.71

to

0.70

 viewpoint

0.67

Activations Density 0.499%

numbers and numerical progressions

numerical values related to measurements and statistics

No Comments

No Known Activations